I attended Vinod Khosla’s talk in Bangalore a few days back. His topic was Solar Thermal Power (which involves power generation using steam made in a solar contraption). Luckily his talk ended being broader than that. He talked about ethanol, venture investing, entrepreneurship and eradicating poverty in India. Rajan has a short summary of the discussion here.
I know only a little about renewable energy. I had heard about the raging debate on ethanol from the Scientific American article in January (subscription required). It’s only after Vinod’s talk though that I realized that this is a big issue and there is much passion on both sides. The arguments for and against ethanol are quite interesting. But that’s not the Big Fight that I want to talk about.
The Big Fight
The Big Fight that I am talking about is the one between BigCo’s and venture-funded NewCos in the energy space. So far the energy sector has been dominated by the BigCo’s. After all, who other than a BigCo, say GE, can afford to invest billions of dollars into a next-gen nuclear power plant and then wait 10 years for results? Apart from nuclear power, solar cells, oil exploration, even hybrid cars have all been the domain of BigCos. But now venture-backed NewCo’s are moving in. Economist suggests that over $2b of venture capital was invested in this area in 2006 alone.
A Big Fight has begun in earnest in the energy space.
I have suggested earlier that the BigCo innovation model is quite different from the venture-backed NewCo model. BigCos use the kaizen approach of continuous incremental innovation. They display patience and an appetite for big capital investments. In contrast, the NewCos innovation model is based on a swarm approach. Many try, each experimenting with a different strategy. Most fail, but a few succeed. And a new pathway is created that opens up a new world of opportunities.
I have seen this battle of innovation ideologies play out earlier. Quite some time back I wrote a post, War of Innovation Ideologies, where I talked about the telecom switch market…
In the early-80s there were only 3-4 companies worldwide that had the wherewithal to invest in creating a digital telephone switch. But by the late-90s, the capital requirements had reduced and the new generation telephone switch, now called softswitch, could be built on standard high-performance computer platforms. Not surprisingly startups became the innovation leaders in this category.
Nowadays there is the same battle about to play out for radios in mobile networks (called base transceiver stations). The next generation products will use software-defined radios built on mostly off-the-shelf platforms. It’s a safe bet that startups will set the pace of innovation in this space.
Regular readers know that I am deeply interested in the same battle playing out in the pharma/biotech space.
In each of these cases I have been on the side of the venture-funded NewCos. Yet I can’t get myself to take the same position for the energy sector. What gives?
Missing in action
I believe that for venture-backed NewCo’s to win the Big Fight, there has to be a major reduction in capital intensity in that sector. This can either happen due to new technology or can be a result of subtle but significant value chain evolution. Without a dramatic reduction in capital intensity, legions of startups cannot be setup to attack the problem(s) in that sector. And unless there is intense experimentation taking place through these legions of startups, the garage model doesn’t do better than the kaizen model of innovation. This is the hypothesis that I have. I admit that it’s not tested statistically but it lines up with every situation that I know.
Coming back to the energy sector, I don’t yet see a reason for a dramatic reduction in capital intensity. There isn’t new technology that’s come to the market yet (as was the case for telephone switches or is now the case for software-defined radios). Nor is there any evidence that significant value chain evolution has taken place (as was the case for fabless chip design shops, or might be the case for pharma/biotech sector).
Obviously things might change in a few years as university research picks up and some technology breakthroughs are made. Till that happens it’s quite likely that the gains will go to ADM and GE rather than the startups that Khosla Ventures and others have backed.
Vanity Capital
Don’t misunderstand me here; there is nothing wrong in making these bets. But keep in mind that this is vanity capital investing not traditional venture investing. It’s like the investments that Richard Branson and Jeff Bezos have made into the space travel sector. Surely they are making a difference. Yet it’s unclear as to where the value capture will take place – in Boeing/Lockheed or in startups.
I guess if I was to make this argument to Vinod Khosla, he would say in his defense that he is not focusing on solar cells; instead he is investing in solar thermals. Or that he is not focusing on oil exploration but on ethanol production. Or even that he is not investing in hybrid cars or batteries but in small fuel cells. In other words, he is being discriminating about where the BigCo’s will win and where the NewCo’s will win. I am not convinced. It sure looks like vanity capital investing to me right now. If I was a VC I would wait a few years before getting in. What do you think?
Thanks for the trackback, long post (Hagel Style
) but a good post 
. Agree with your overall point but I don’t think I agree with your example of software radio. From what I have seen SDR has to still go a very long way for the capital intensity to reduce , I have not seen good workable products come out yet, Vanu is the best in the league and they have made great strides but still not good enough in the context in which you are speaking. High end ADC is still the bottleneck.
- Rajan
Rajan - Software defined radios have Moore’s law working in its favor. They are about to become visible in the consumer space. On the BTS side, Vanu has over 30 of them in trials. I am aware of at least one other stealth mode startup in this space. There is lots happening in this space.
Sharad, To your swarm innovation point - there was a sea change in the way big companies are approaching innovation around the early 90s. Instead of running a research lab to innovate, now they would rather let the market sponsor innovation (venture capital?) and then purchase the most successful experiment.
Could the energy companies not end up in a similar situation? Newco may never be a successful business but could be a successful innovation that a bigco ends up adopting.
Hi Ashish – You are right that swarm model of innovation requires a symbiotic relationship between BigCos and NewCos. But for the swarm to form, the cost of individual experimentation (i.e. capital intensity) has to come down dramatically. The reason that a swarm could never have formed for digital telephone switches is simply because each experiment costed about a billion dollars (in today’s terms). So only a few BigCos could afford this. My core point is that high capital intensity is the bane of nuclear power, solar cells (materials research is very expensive), ethanol pathways, etc. So either BigCos have to fund the few “experiments” taking place or the universities have to step in to do so. So far the university spending in renewable energy has been pathetic, particularly in US. This is why I feel that the big problems in the energy space haven’t yet become open to to the swarm model of innovation.
There are inevitably a few smaller areas, perhaps like solar thermal solutions, where the capital intensity is low enough for the swarm model to work. There, as you point out, the winners will get bought out by the BigCos. The question for me is: at what valuation? I haven’t addressed this in the original post so let me take quick shot at this.
A BigCo’s buy-vs-build analysis works in favor of startups when one of two conditions are met. First, the startup should have done things in a way that it’s much cheaper to buy than build. This happens as much due to typical startup frugality as it does because of its willingness to exploit the technology curve to its advantage. If there technology improvement curve itself is pretty flat, as is the case in the energy sector right now, getting a dramatic saving from buy-decision are harder to obtain.
The second condition that has to be met is to do with time-to-market. It may not be cheaper to buy a startup but that may be the only way to hit the market window. In that respect, I fear that the energy sector is like mobile 3G market; there isn’t enough market hustle to make a change happen quickly. Despite the big sums of money that the mobile operators paid for 3G licenses, they were not in a hurry to deploy 3G networks (as they wanted to milk their investments in 2.5G first). In most markets, the deployments promised for 2002 haven’t still happened. Since there was no rush, there was little incentive for BigCos (Ericsson, Lucent, etc) to buy rather than build. That’s why most of 3G network element startups never saw meaningful exits.
But energy sector is different, you might say. After all, there is market pressure to move forward and change the status-quo. I would caution you there. I grant you that there is consensus that the current status-quo is broken but that doesn’t mean that whole value-chain is about to change. Look at US healthcare, outside of drug discovery, in areas like patient record management, etc. an example. Everybody concedes that the current system is broken and needs a fix. But NewCos doing good work aren’t getting snapped up yet. It’s not a favorable space for startups.
The bottom-line of all this is about timing. It seems to me that we are still in the vanity investing stage stage of market evolution. Soon, perhaps a few years from now, vanity investing (along with BigCo investing and university research) would have opened up this sector for venture investing stage. That would be the time to get in. This is just a humble opinion.
Thanks for commenting! Hope to see you here more often.